At present, the error control method for high-speed serial data transmission obtains the errors by comparison and then controls them. If the data transmission channel is not denoised, the packet loss and error codes become serious, and energy consumption increases. The use of fuzzy classification is proposed to control data transmission errors. The method uses the combination of wavelet transform and transform domain difference to double denoise the channel, and it completes the clustering of data transmission errors by fuzzy classification. Considering packet loss, error codes, and energy consumption in data transmission error control, when the communication distance between two nodes is small, automatic repeat request is used to control data transmission errors. As the distance between nodes increases, forward error correction is used to control data transmission errors. When the communication distance gradually increases, data transmission errors are controlled by hybrid automatic repeat request. Experiments showed that the proposed method can reduce the data transmission error, control energy consumption, packet loss rate, and bit error rate, and enhance the denoising effect.